用模糊人工神经网络方法预测多层砖房震害

结构的震害预测是编制抗震防灾规划的重要基础内容。应用模糊人工神经网络方法来对结构震害进行分析预测。应针对不同类型的构造物筛选出影响其震害的主要因素 ,采用地区性的历史震害资料对所建成的人工神经网络进行训练的方法。以量大面广的多层砖房为例 ,以云南省近几年来的几次大地震的灾害资料为样本 ,运用灰色数学方法筛选出影响震害的主要因素 ,然后进行了模糊人工神经网络的构筑与训练。结果表明 ,这种方法的训练效果是令人满意的

Forecasting seismic damage in multistory masonry buildings with a neuro-fuzzy approach

A fuzzy artificial neural networks is used to analyze and forecast seismic damage in multistory masonry buildings. Different types of structures will incur very different seismic damage, even in the same location. Thus, the seismic damage influence factors must include the structure type. Gray system method was used to select the major seismic damage influence factors. Local seismic damage history data used to train the artificial neural network was for multistory masonry buildings, which have been built in many locations in Yunnan, China. The results demonstrate the efficiency of the method.